SNAP Data Extraction and Bias Correction File

Setting up the data so that the difference between the SNAP data and the airport specific data.

Step 1: Extract the RCP Site data from the SNAP 5 model averages.

Step 2: Import both data sets

Step 3: Set up the Z-score calculation based on notes from meeting with Alan Hamlet

Step 4: Calculate the average and standard deviation of the daily differences

Step 5 (not in this file): rerun models with the new forecast temperatures – List these files here after the fact to make sure this is up to date

Load in the packages needed

library(dplyr)

Setting up the workspace

library(dplyr)
library(ggplot2)

setwd("~/Library/CloudStorage/Dropbox/SWEL/CopperRiverDelta/PondTemps/DataAnalysis/PondTempsAnalysis")

Extracting the RCP data from the aat file

Read in the data

load("~/Dropbox/SWEL/CopperRiverDelta/PondTemps/DataAnalysis/PondTempsAnalysis/forecasted_temperature.RData")

str(aat)

'data.frame':   32040 obs. of  7 variables:
 $ temperature: num  -7.1 -6.3 0 -8.3 -4.9 -3.4 -6.8 -12.5 -2.7 -6 ...
 $ rcp        : chr  "45" "45" "45" "45" ...
 $ model      : chr  "GFDL-CM3" "GFDL-CM3" "GFDL-CM3" "GFDL-CM3" ...
 $ month      : chr  "01" "01" "01" "01" ...
 $ year       : chr  "2012" "2013" "2014" "2015" ...
 $ site       : chr  "yak" "yak" "yak" "yak" ...
 $ date       : 'yearmon' num  Jan 2012 Jan 2013 Jan 2014 Jan 2015 ...

This comes in as a datafile with columns temperature, rcp, model, month, year, site, and date

Extract the data

# selecting CRD
CRD <- aat %>%
  dplyr::filter(site == "cord")

# filter for each rcp
CR_4.5 <- CRD %>%
  dplyr::filter(rcp == "45")

CR_6.0 <- CRD %>%
  dplyr::filter(rcp == "60")

CR_8.5 <- CRD %>%
  dplyr::filter(rcp == "85")

# selecting YF
YF <- aat %>%
  dplyr::filter(site == "yak")

# filter for each rcp
YF_4.5 <- YF %>%
  dplyr::filter(rcp == "45")

YF_6.0 <- YF %>%
  dplyr::filter(rcp == "60")

YF_8.5 <- YF%>%
  dplyr::filter(rcp == "85")

ggplot()+
   geom_smooth(data = CR_8.5, aes(x = date, y = temperature, color = "red")) +
   geom_smooth(data = CR_4.5, aes(x = date, y = temperature, color = "blue"))

`geom_smooth()` using method = 'gam' and formula = 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula = 'y ~ s(x, bs = "cs")'

NA
NA
NA

Starting with the RCP 4.5 data

Filtering to 4.5

CR_4.5

# Group by date (year and month combination) and calculate average temperature
avg_CR_4.5 <- CR_4.5 %>%
  group_by(date) %>%
  summarise(avg_temperature = mean(temperature, na.rm = TRUE),
            std_temperature = sd(temperature, na.rm = TRUE),
            precision = 1 / (sd(temperature, na.rm = TRUE)^2))

# View the resulting dataframe
print(avg_CR_4.5)

CR_4.5 <- avg_CR_4.5
CR_4.5

YF_4.5

# Group by date (year and month combination) and calculate average temperature
avg_YF_4.5 <- YF_4.5 %>%
  group_by(date) %>%
  summarise(avg_temperature = mean(temperature, na.rm = TRUE),
            std_temperature = sd(temperature, na.rm = TRUE),
            precision = 1 / (sd(temperature, na.rm = TRUE)^2))

# View the resulting dataframe
print(avg_YF_4.5)

YF_4.5 <- avg_YF_4.5
YF_4.5

Plot the forecasted data

CR_4.5

# Full time frame (2006-2100)
CR_SNAP <- ggplot(data = CR_4.5, mapping = aes(x = date)) +
  geom_line(aes(y = avg_temperature), size = 0.8) +  # Line plot for avg_temperature
  geom_ribbon(aes(ymin = avg_temperature - std_temperature, 
                  ymax = avg_temperature + std_temperature),  # Ribbon for uncertainty
              alpha = 0.5, fill = "blue") +
  xlab("Year") +
  ylab("Mean Forecast Temperature") +
  ggtitle("Copper River SNAP Forecast - RCP 4.5") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5))
CR_SNAP


ggsave("CR_Geo_Forecast4.5_2020.jpeg", plot = CR_SNAP, width = 8, height = 6)

YF_4.5

# Full time frame (2006-2100)
YF_SNAP <- ggplot(data = YF_4.5, mapping = aes(x = date)) +
  geom_line(aes(y = avg_temperature), size = 0.8) +  # Line plot for avg_temperature
  geom_ribbon(aes(ymin = avg_temperature - std_temperature,
                  ymax = avg_temperature + std_temperature),  # Ribbon for uncertainty
              alpha = 0.5, fill = "blue") +
  xlab("Year") +
  ylab("Mean Forecast Temperature") +
  ggtitle("Yakutat Foreland SNAP Forecast - RCP 4.5") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5))
YF_SNAP


ggsave("YF_Geo_Forecast4.5_2020.jpeg", plot = YF_SNAP, width = 8, height = 6)

Continuing with the RCP 6.0 data

Filtering to 6.0

CR_6.0

# Group by date (year and month combination) and calculate average temperature
avg_CR_6.0 <- CR_6.0 %>%
  group_by(date) %>%
  summarise(avg_temperature = mean(temperature, na.rm = TRUE),
            std_temperature = sd(temperature, na.rm = TRUE),
            precision = 1 / (sd(temperature, na.rm = TRUE)^2))

# View the resulting dataframe
print(avg_CR_6.0)

CR_6.0 <- avg_CR_6.0
CR_6.0

YF_6.0

# Group by date (year and month combination) and calculate average temperature
avg_YF_6.0 <- YF_6.0 %>%
  group_by(date) %>%
  summarise(avg_temperature = mean(temperature, na.rm = TRUE),
            std_temperature = sd(temperature, na.rm = TRUE),
            precision = 1 / (sd(temperature, na.rm = TRUE)^2))

# View the resulting dataframe
print(avg_YF_6.0)

YF_6.0 <- avg_YF_6.0
YF_6.0

Plot the forecasted data

CR_6.0

# Full time frame (2006-2100)
CR_SNAP <- ggplot(data = CR_6.0, mapping = aes(x = date)) +
  geom_line(aes(y = avg_temperature), size = 0.8) +  # Line plot for avg_temperature
  geom_ribbon(aes(ymin = avg_temperature - std_temperature, 
                  ymax = avg_temperature + std_temperature),  # Ribbon for uncertainty
              alpha = 0.5, fill = "blue") +
  xlab("Year") +
  ylab("Mean Forecast Temperature") +
  ggtitle("Copper River SNAP Forecast - RCP 6.0") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5))
CR_SNAP


ggsave("CR_Geo_Forecast6.0_2020.jpeg", plot = CR_SNAP, width = 8, height = 6)

YF_6.0

# Full time frame (2006-2100)
YF_SNAP <- ggplot(data = YF_6.0, mapping = aes(x = date)) +
  geom_line(aes(y = avg_temperature), size = 0.8) +  # Line plot for avg_temperature
  geom_ribbon(aes(ymin = avg_temperature - std_temperature,
                  ymax = avg_temperature + std_temperature),  # Ribbon for uncertainty
              alpha = 0.5, fill = "blue") +
  xlab("Year") +
  ylab("Mean Forecast Temperature") +
  ggtitle("Yakutat Foreland SNAP Forecast - RCP 6.0") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5))
YF_SNAP


ggsave("YF_Geo_Forecast6.0_2020.jpeg", plot = YF_SNAP, width = 8, height = 6)

Continuing with the RCP 8.5 data

Filtering to 8.5

CR_8.5

# Group by date (year and month combination) and calculate average temperature
avg_CR_8.5 <- CR_8.5 %>%
  group_by(date) %>%
  summarise(avg_temperature = mean(temperature, na.rm = TRUE),
            std_temperature = sd(temperature, na.rm = TRUE),
            precision = 1 / (sd(temperature, na.rm = TRUE)^2))

# View the resulting dataframe
print(avg_CR_8.5)

CR_8.5 <- avg_CR_8.5
CR_8.5

YF_8.5

# Group by date (year and month combination) and calculate average temperature
avg_YF_8.5 <- YF_8.5 %>%
  group_by(date) %>%
  summarise(avg_temperature = mean(temperature, na.rm = TRUE),
            std_temperature = sd(temperature, na.rm = TRUE),
            precision = 1 / (sd(temperature, na.rm = TRUE)^2))

# View the resulting dataframe
print(avg_YF_8.5)

YF_8.5 <- avg_YF_8.5
YF_8.5

Plot the forecasted data

CR_8.5

# Full time frame (2006-2100)
CR_SNAP <- ggplot(data = CR_8.5, mapping = aes(x = date)) +
  geom_line(aes(y = avg_temperature), size = 0.8) +  # Line plot for avg_temperature
  geom_ribbon(aes(ymin = avg_temperature - std_temperature, 
                  ymax = avg_temperature + std_temperature),  # Ribbon for uncertainty
              alpha = 0.5, fill = "blue") +
  xlab("Year") +
  ylab("Mean Forecast Temperature") +
  ggtitle("Copper River SNAP Forecast - RCP 8.5") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5))
CR_SNAP


ggsave("CR_Geo_Forecast8.5_2020.jpeg", plot = CR_SNAP, width = 8, height = 6)

YF_8.5

# Full time frame (2006-2100)
YF_SNAP <- ggplot(data = YF_8.5, mapping = aes(x = date)) +
  geom_line(aes(y = avg_temperature), size = 0.8) +  # Line plot for avg_temperature
  geom_ribbon(aes(ymin = avg_temperature - std_temperature,
                  ymax = avg_temperature + std_temperature),  # Ribbon for uncertainty
              alpha = 0.5, fill = "blue") +
  xlab("Year") +
  ylab("Mean Forecast Temperature") +
  ggtitle("Yakutat Foreland SNAP Forecast - RCP 8.5") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5))
YF_SNAP


ggsave("YF_Geo_Forecast8.5_2020.jpeg", plot = YF_SNAP, width = 8, height = 6)

Saving the data as datafiles for later use

write.csv(YF_4.5, file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF45.csv", row.names = FALSE)
write.csv(YF_6.0, file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF60.csv", row.names = FALSE)
write.csv(YF_8.5, file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF85.csv", row.names = FALSE)

write.csv(CR_4.5, file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR45.csv", row.names = FALSE)
write.csv(CR_6.0, file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR60.csv", row.names = FALSE)
write.csv(CR_8.5, file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR85.csv", row.names = FALSE)

If these above is already done, start here and just read in the files that were created ^

YF_45 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF45.csv")
YF_85 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF85.csv")

CR_45 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR45.csv")
CR_85 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR85.csv")

Read in the data file for the airport temperature data Separate this by airport location

CRD_airport <- read.csv(file = "BiasCorrection/CR_PondAirTemp_mean.csv")
CRD_airport <- CRD_airport %>%
  select(-X)

YF_airport <- read.csv(file = "BiasCorrection/YF_PondAirTemp_mean.csv")
YF_airport <- YF_airport %>%
  select(-X)

Copper River RCP 4.5

Read in the data

# Read data files
CR_45 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR45.csv")
CRD_airport <- read.csv(file = "BiasCorrection/CR_PondAirTemp_mean.csv")
CRD_airport <- CRD_airport %>% select(-X)  # Remove unnecessary column 'X'

Check dates

# Fix 'date' in CR_45 to ensure it's in Date format
CR_45$date <- as.Date(paste("01", CR_45$date), format = "%d %b %Y")

# Ensure that there are no NAs in the 'date' column after conversion
sum(is.na(CR_45$date))  # Check if there are any missing values (should be 0)

[1] 0

Trim dates

# Trim CR_45 to the date range of CRD_airport
start_date <- min(CRD_airport$Date)
end_date <- max(CRD_airport$Date)

# Subset CR_45 to match the date range of CRD_airport
CR_45_trimmed <- CR_45 %>%
  filter(date >= start_date & date <= end_date)

Plot pre-correction data

library(ggplot2)

# Ensure 'Date' in CRD_airport is in Date format
CRD_airport$Date <- as.Date(CRD_airport$Date)

# Check structure of CR_45_trimmed to confirm correct Date column
str(CR_45_trimmed)

'data.frame':   96 obs. of  4 variables:
 $ date           : Date, format: "2012-01-01" "2012-02-01" "2012-03-01" ...
 $ avg_temperature: num  -6.66 -3.84 -0.02 3.28 7.2 ...
 $ std_temperature: num  1.483 5.188 2.846 0.944 0.728 ...
 $ precision      : num  0.455 0.0371 0.1235 1.1211 1.8868 ...

# Plot Observed vs Uncorrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = CR_45_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2) +
  geom_line(data = CRD_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Uncorrected Predicted Temperatures - CRD RCP 4.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Pre-CorrectionComparison_CR45.jpeg", plot = Plot, width = 8, height = 6)

z-score bias correction

# Calculate mean and standard deviation for observed and predicted temperatures
mu_obs <- mean(CRD_airport$Air_MonthAvg, na.rm = TRUE)
sigma_obs <- sd(CRD_airport$Air_MonthAvg, na.rm = TRUE)

mu_pred <- mean(CR_45_trimmed$avg_temperature, na.rm = TRUE)
sigma_pred <- sd(CR_45_trimmed$avg_temperature, na.rm = TRUE)

Correction applied to the 2012-2020 dataset

# Apply the z-score bias correction to the predicted data
CR_45_trimmed$corrected_temp <- (CR_45_trimmed$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(CR_45_trimmed)

NA

Plot corrected and observed values

# Plotting Observed vs Corrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = CR_45_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  geom_line(data = CRD_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Corrected Predicted Temperatures - CRD RCP 4.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Corrected)" = "blue", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    plot.subtitle = element_text(size = 12, face = "italic"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Corrected_CR45.jpeg", plot = Plot, width = 8, height = 6)

Plotting the values here

# Plot Original vs Corrected Predicted Temperatures (No Observed Values)
Plot <- ggplot() +
  # Plot the original predicted temperatures (uncorrected)
  geom_line(data = CR_45_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2, linetype = "solid") +
  
  # Plot the corrected predicted temperatures
  geom_line(data = CR_45_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  
  # Set plot labels
  labs(
    title = "Original vs Corrected Predicted Temperatures - CRD RCP 4.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  
  # Customize x-axis to show months/years clearly
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  
  # Define custom colors for the plot
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Predicted (Corrected)" = "blue")) +
  
  # Minimal theme for the plot
  theme_minimal() +
  
  # Customizations for appearance
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),  # Rotate x-axis labels for clarity
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/CorrectedComparison_CR45.jpeg", plot = Plot, width = 8, height = 6)

Correction applied to the whole dataset

# Apply the z-score bias correction to the predicted data
CR_45$corrected_temp <- (CR_45$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(CR_45)

Plot this to make it into the same predicted (corrected) + predicted (uncorrected) with the error included

# Plot Original vs Corrected Predicted Temperatures (No Observed Values)
Plot <- ggplot() +
  # Plot the original predicted temperatures (uncorrected)
  geom_line(data = CR_45, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2, linetype = "solid") +
  
  # Plot the corrected predicted temperatures
  geom_line(data = CR_45, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  
  geom_line(data = CRD_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  
  # Set plot labels
  labs(
    title = "Original vs Corrected Predicted Temperatures - CRD RCP 4.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  
  # Customize x-axis to show months/years clearly
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  
  # Define custom colors for the plot
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Predicted (Corrected)" = "blue")) +
  
  # Minimal theme for the plot
  theme_minimal() +
  
  # Customizations for appearance
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),  # Rotate x-axis labels for clarity
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/FullCorrection_CR45.jpeg", plot = Plot, width = 8, height = 6)

Save the new file

# Save the updated dataset with the corrected temperatures to a new CSV file
write.csv(CR_45, "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR45_corrected.csv", row.names = FALSE)

Copper River RCP 8.5

Read in the data

# Read data files
CR_85 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR85.csv")
CRD_airport <- read.csv(file = "BiasCorrection/CR_PondAirTemp_mean.csv")
CRD_airport <- CRD_airport %>% select(-X)  # Remove unnecessary column 'X'

Check dates

# Fix 'date' in CR_85 to ensure it's in Date format
CR_85$date <- as.Date(paste("01", CR_85$date), format = "%d %b %Y")

# Ensure that there are no NAs in the 'date' column after conversion
sum(is.na(CR_85$date))  # Check if there are any missing values (should be 0)

[1] 0

Trim dates

# Trim CR_85 to the date range of CRD_airport
start_date <- min(CRD_airport$Date)
end_date <- max(CRD_airport$Date)

# Subset CR_85 to match the date range of CRD_airport
CR_85_trimmed <- CR_85 %>%
  filter(date >= start_date & date <= end_date)

Plot pre-correction data

library(ggplot2)

# Ensure 'Date' in CRD_airport is in Date format
CRD_airport$Date <- as.Date(CRD_airport$Date)

# Check structure of CR_85_trimmed to confirm correct Date column
str(CR_85_trimmed)

'data.frame':   96 obs. of  4 variables:
 $ date           : Date, format: "2012-01-01" "2012-02-01" "2012-03-01" ...
 $ avg_temperature: num  -6.54 -4.5 -1.92 3.24 6.64 ...
 $ std_temperature: num  4.46 4.2 3.89 1.23 1.12 ...
 $ precision      : num  0.0503 0.0568 0.0662 0.6631 0.7981 ...

# Plot Observed vs Uncorrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = CR_85_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2) +
  geom_line(data = CRD_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Uncorrected Predicted Temperatures - CRD RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Pre-CorrectionComparison_CR85.jpeg", plot = Plot, width = 8, height = 6)

z-score bias correction

# Calculate mean and standard deviation for observed and predicted temperatures
mu_obs <- mean(CRD_airport$Air_MonthAvg, na.rm = TRUE)
sigma_obs <- sd(CRD_airport$Air_MonthAvg, na.rm = TRUE)

mu_pred <- mean(CR_85_trimmed$avg_temperature, na.rm = TRUE)
sigma_pred <- sd(CR_85_trimmed$avg_temperature, na.rm = TRUE)

Correction applied to the 2012-2020 dataset

# Apply the z-score bias correction to the predicted data
CR_85_trimmed$corrected_temp <- (CR_85_trimmed$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(CR_85_trimmed)

NA

Plot corrected and observed values

# Plotting Observed vs Corrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = CR_85_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  geom_line(data = CRD_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Corrected Predicted Temperatures - CRD RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Corrected)" = "blue", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    plot.subtitle = element_text(size = 12, face = "italic"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Corrected_CR85.jpeg", plot = Plot, width = 8, height = 6)

Plotting the values here

# Plot Original vs Corrected Predicted Temperatures (No Observed Values)
Plot <- ggplot() +
  # Plot the original predicted temperatures (uncorrected)
  geom_line(data = CR_85_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2, linetype = "solid") +
  
  # Plot the corrected predicted temperatures
  geom_line(data = CR_85_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  
  # Set plot labels
  labs(
    title = "Original vs Corrected Predicted Temperatures - CRD RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  
  # Customize x-axis to show months/years clearly
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  
  # Define custom colors for the plot
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Predicted (Corrected)" = "blue")) +
  
  # Minimal theme for the plot
  theme_minimal() +
  
  # Customizations for appearance
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),  # Rotate x-axis labels for clarity
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/CorrectedComparison_CR85.jpeg", plot = Plot, width = 8, height = 6)

Correction applied to the whole dataset

# Apply the z-score bias correction to the predicted data
CR_85$corrected_temp <- (CR_85$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(CR_85)

Plot this to make it into the same predicted (corrected) + predicted (uncorrected) with the error included

# Plot Original vs Corrected Predicted Temperatures (No Observed Values)
Plot <- ggplot() +
  # Plot the original predicted temperatures (uncorrected)
  geom_line(data = CR_85, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2, linetype = "solid") +
  
  # Plot the corrected predicted temperatures
  geom_line(data = CR_85, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  
  geom_line(data = CRD_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  
  # Set plot labels
  labs(
    title = "Original vs Corrected Predicted Temperatures - CRD RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  
  # Customize x-axis to show months/years clearly
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  
  # Define custom colors for the plot
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Predicted (Corrected)" = "blue")) +
  
  # Minimal theme for the plot
  theme_minimal() +
  
  # Customizations for appearance
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),  # Rotate x-axis labels for clarity
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/FullCorrection_CR85.jpeg", plot = Plot, width = 8, height = 6)

Save the new file

# Save the updated dataset with the corrected temperatures to a new CSV file
write.csv(CR_85, "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_CR85_corrected.csv", row.names = FALSE)

Yakutat Foreland RCP 4.5

Read in the data

# Read data files
YF_45 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF45.csv")
YF_airport <- read.csv(file = "BiasCorrection/YF_PondAirTemp_mean.csv")
YF_airport <- YF_airport %>% select(-X)  # Remove unnecessary column 'X'

Check dates

# Fix 'date' in YF_45 to ensure it's in Date format
YF_45$date <- as.Date(paste("01", YF_45$date), format = "%d %b %Y")

# Ensure that there are no NAs in the 'date' column after conversion
sum(is.na(YF_45$date))  # Check if there are any missing values (should be 0)

[1] 0

Trim dates

# Trim YF_45 to the date range of YF_airport
start_date <- min(YF_airport$Date)
end_date <- max(YF_airport$Date)

# Subset YF_45 to match the date range of YF_airport
YF_45_trimmed <- YF_45 %>%
  filter(date >= start_date & date <= end_date)

Plot pre-correction data

library(ggplot2)

# Ensure 'Date' in YF_airport is in Date format
YF_airport$Date <- as.Date(YF_airport$Date)

# Check structure of YF_45_trimmed to confirm correct Date column
str(YF_45_trimmed)

'data.frame':   96 obs. of  4 variables:
 $ date           : Date, format: "2012-01-01" "2012-02-01" "2012-03-01" ...
 $ avg_temperature: num  -6.76 -3.42 0.04 3.58 7.46 ...
 $ std_temperature: num  0.695 5.172 2.487 0.986 0.893 ...
 $ precision      : num  2.0704 0.0374 0.1617 1.0288 1.2531 ...

# Plot Observed vs Uncorrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = YF_45_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2) +
  geom_line(data = YF_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Uncorrected Predicted Temperatures - YF 4.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Pre-CorrectionComparison_YF45.jpeg", plot = Plot, width = 8, height = 6)

z-score bias correction

# Calculate mean and standard deviation for observed and predicted temperatures
mu_obs <- mean(YF_airport$Air_MonthAvg, na.rm = TRUE)
sigma_obs <- sd(YF_airport$Air_MonthAvg, na.rm = TRUE)

mu_pred <- mean(YF_45_trimmed$avg_temperature, na.rm = TRUE)
sigma_pred <- sd(YF_45_trimmed$avg_temperature, na.rm = TRUE)

Correction applied to the 2012-2020 dataset

# Apply the z-score bias correction to the predicted data
YF_45_trimmed$corrected_temp <- (YF_45_trimmed$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(YF_45_trimmed)

NA

Plot corrected and observed values

# Plotting Observed vs Corrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = YF_45_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  geom_line(data = YF_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Corrected Predicted Temperatures - YF RCP 4.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Corrected)" = "blue", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    plot.subtitle = element_text(size = 12, face = "italic"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Corrected_YF45.jpeg", plot = Plot, width = 8, height = 6)

Plotting the values here

<!-- rnb-source-end -->

<!-- rnb-output-begin eyJkYXRhIjoiRXJyb3I6IGF0dGVtcHQgdG8gdXNlIHplcm8tbGVuZ3RoIHZhcmlhYmxlIG5hbWVcbiJ9 -->

Error: attempt to use zero-length variable name

<!-- rnb-output-end -->

<!-- rnb-chunk-end -->


<!-- rnb-text-begin -->

Correction applied to the whole dataset


<!-- rnb-text-end -->


<!-- rnb-chunk-begin -->


<!-- rnb-frame-begin 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 -->

<div data-pagedtable="false">
  <script data-pagedtable-source type="application/json">
{"columns":[{"label":[""],"name":["_rn_"],"type":[""],"align":["left"]},{"label":["date"],"name":[1],"type":["date"],"align":["right"]},{"label":["avg_temperature"],"name":[2],"type":["dbl"],"align":["right"]},{"label":["std_temperature"],"name":[3],"type":["dbl"],"align":["right"]},{"label":["precision"],"name":[4],"type":["dbl"],"align":["right"]},{"label":["corrected_temp"],"name":[5],"type":["dbl"],"align":["right"]}],"data":[{"1":"2012-01-01","2":"-6.76","3":"0.6949820","4":"2.07039337","5":"-4.092163","_rn_":"1"},{"1":"2012-02-01","2":"-3.42","3":"5.1722336","4":"0.03738038","5":"-1.251206","_rn_":"2"},{"1":"2012-03-01","2":"0.04","3":"2.4865639","4":"0.16173379","5":"1.691821","_rn_":"3"},{"1":"2012-04-01","2":"3.58","3":"0.9859006","4":"1.02880658","5":"4.702895","_rn_":"4"},{"1":"2012-05-01","2":"7.46","3":"0.8933085","4":"1.25313283","5":"8.003168","_rn_":"5"},{"1":"2012-06-01","2":"12.72","3":"2.2487774","4":"0.19774570","5":"12.477249","_rn_":"6"}],"options":{"columns":{"min":{},"max":[10],"total":[5]},"rows":{"min":[10],"max":[10],"total":[6]},"pages":{}}}
  </script>
</div>

<!-- rnb-frame-end -->

<!-- rnb-source-begin eyJkYXRhIjoiYGBgclxuIyBDaGVjayB0aGUgZmlyc3QgZmV3IHJvd3MgdG8gdmVyaWZ5XG5oZWFkKFlGXzQ1KVxuXG5cbmBgYCJ9 -->

```r
# Check the first few rows to verify
head(YF_45)

Plot this to make it into the same predicted (corrected) + predicted (uncorrected) with the error included

Save the new file

# Save the updated dataset with the corrected temperatures to a new CSV file
write.csv(YF_45, "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF45_corrected.csv", row.names = FALSE)

Yakutat Foreland RCP 8.5

Read in the data

# Read data files
YF_85 <- read.csv(file = "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF85.csv")
YF_airport <- read.csv(file = "BiasCorrection/YF_PondAirTemp_mean.csv")
YF_airport <- YF_airport %>% select(-X)  # Remove unnecessary column 'X'

Check dates

# Fix 'date' in YF_85 to ensure it's in Date format
YF_85$date <- as.Date(paste("01", YF_85$date), format = "%d %b %Y")

# Ensure that there are no NAs in the 'date' column after conversion
sum(is.na(YF_85$date))  # Check if there are any missing values (should be 0)

[1] 0

Trim dates

# Trim YF_85 to the date range of YF_airport
start_date <- min(YF_airport$Date)
end_date <- max(YF_airport$Date)

# Subset YF_85 to match the date range of YF_airport
YF_85_trimmed <- YF_85 %>%
  filter(date >= start_date & date <= end_date)

Plot pre-correction data

library(ggplot2)

# Ensure 'Date' in YF_airport is in Date format
YF_airport$Date <- as.Date(YF_airport$Date)

# Check structure of YF_85_trimmed to confirm correct Date column
str(YF_85_trimmed)

'data.frame':   96 obs. of  4 variables:
 $ date           : Date, format: "2012-01-01" "2012-02-01" "2012-03-01" ...
 $ avg_temperature: num  -5.88 -4.18 -1.46 3.66 6.7 ...
 $ std_temperature: num  3.528 3.283 3.769 1.537 0.987 ...
 $ precision      : num  0.0803 0.0928 0.0704 0.4232 1.0256 ...

# Plot Observed vs Uncorrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = YF_85_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2) +
  geom_line(data = YF_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Uncorrected Predicted Temperatures - YF RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Pre-CorrectionComparison_YF85.jpeg", plot = Plot, width = 8, height = 6)

z-score bias correction

# Calculate mean and standard deviation for observed and predicted temperatures
mu_obs <- mean(YF_airport$Air_MonthAvg, na.rm = TRUE)
sigma_obs <- sd(YF_airport$Air_MonthAvg, na.rm = TRUE)

mu_pred <- mean(YF_85_trimmed$avg_temperature, na.rm = TRUE)
sigma_pred <- sd(YF_85_trimmed$avg_temperature, na.rm = TRUE)

Correction applied to the 2012-2020 dataset

# Apply the z-score bias correction to the predicted data
YF_85_trimmed$corrected_temp <- (YF_85_trimmed$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(YF_85_trimmed)

NA

Plot corrected and observed values

# Plotting Observed vs Corrected Predicted Temperatures
Plot <- ggplot() +
  geom_line(data = YF_85_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  geom_line(data = YF_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  labs(
    title = "Observed vs Corrected Predicted Temperatures - YF RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  scale_color_manual(values = c("Predicted (Corrected)" = "blue", "Observed" = "black")) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    plot.subtitle = element_text(size = 12, face = "italic"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/Corrected_YF85.jpeg", plot = Plot, width = 8, height = 6)

Plotting the values here

# Plot Original vs Corrected Predicted Temperatures (No Observed Values)
Plot <- ggplot() +
  # Plot the original predicted temperatures (uncorrected)
  geom_line(data = YF_85_trimmed, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2, linetype = "solid") +
  
  # Plot the corrected predicted temperatures
  geom_line(data = YF_85_trimmed, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  
  # Set plot labels
  labs(
    title = "Original vs Corrected Predicted Temperatures - YF RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  
  # Customize x-axis to show months/years clearly
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  
  # Define custom colors for the plot
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Predicted (Corrected)" = "blue")) +
  
  # Minimal theme for the plot
  theme_minimal() +
  
  # Customizations for appearance
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),  # Rotate x-axis labels for clarity
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/CorrectedComparison_YF85.jpeg", plot = Plot, width = 8, height = 6)

Correction applied to the whole dataset

# Apply the z-score bias correction to the predicted data
YF_85$corrected_temp <- (YF_85$avg_temperature - mu_pred) / sigma_pred * sigma_obs + mu_obs

# Check the first few rows to verify
head(YF_85)

Plot this to make it into the same predicted (corrected) + predicted (uncorrected) with the error included

# Plot Original vs Corrected Predicted Temperatures (No Observed Values)
Plot <- ggplot() +
  # Plot the original predicted temperatures (uncorrected)
  geom_line(data = YF_85, aes(x = date, y = avg_temperature, color = "Predicted (Uncorrected)"), size = 1.2, linetype = "solid") +
  
  # Plot the corrected predicted temperatures
  geom_line(data = YF_85, aes(x = date, y = corrected_temp, color = "Predicted (Corrected)"), size = 1.2, linetype = "dashed") +
  
  geom_line(data = YF_airport, aes(x = Date, y = Air_MonthAvg, color = "Observed"), size = 1.2) +
  
  # Set plot labels
  labs(
    title = "Original vs Corrected Predicted Temperatures - YF RCP 8.5",
    x = "Date",
    y = "Temperature (°C)",
    color = "Temperature Type"
  ) +
  
  # Customize x-axis to show months/years clearly
  scale_x_date(
    date_labels = "%b %Y",    # Format x-axis as Month-Year
    date_breaks = "1 month",  # Set breaks every month
    expand = c(0, 0)          # Remove extra space at the edges
  ) +
  
  # Define custom colors for the plot
  scale_color_manual(values = c("Predicted (Uncorrected)" = "red", "Predicted (Corrected)" = "blue")) +
  
  # Minimal theme for the plot
  theme_minimal() +
  
  # Customizations for appearance
  theme(
    axis.text.x = element_text(angle = 85, hjust = 1),  # Rotate x-axis labels for clarity
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10),
    plot.title = element_text(size = 14, face = "bold"),
    legend.title = element_text(size = 12),
    legend.text = element_text(size = 10)
  )
Plot

ggsave("BiasCorrection/Geo_BiasCorrection/Plots/FullCorrection_YF85.jpeg", plot = Plot, width = 8, height = 6)

Save the new file

# Save the updated dataset with the corrected temperatures to a new CSV file
write.csv(YF_85, "BiasCorrection/Geo_BiasCorrection/DataFiles/SNAP_YF85_corrected.csv", row.names = FALSE)